Documentation
Guides, reference architectures, and control mappings for teams building AI systems that handle protected health data
Guide
Implementation Guide
A six-step path from use-case scoping through production deployment
Assess your AI use case
Pin down the clinical workflow, data types (PHI, imaging, genomics), and which regulations apply (HIPAA, FDA SaMD, state AI laws)
Select compliance pattern
Pick an architecture pattern that matches your regulatory profile: HIPAA-only, GxP, or FDA-regulated
Map security controls
Use the cross-cloud control mappings to identify the specific services you need on AWS, Azure, or GCP
Apply governance protocols
Set up data classification, access policies, lineage tracking, and retention rules for PHI and training data
Run risk assessment
Score AI-specific risks (bias, drift, adversarial inputs, hallucination) alongside standard infrastructure risks
Deploy via playbook
Use a deployment playbook to stand up infrastructure, run validation checks, and promote to production
Resources
Documentation
Architecture Patterns
Reference architectures for HIPAA, GxP, and FDA-regulated AI workloads
- Compliance Automation
- Remote Patient Monitoring
- Serverless Resilience
- Predictive analytics
Security Controls
AWS, Azure, and GCP service mappings for PHI protection, model security, and compliance
- AWS → Azure → GCP equivalents
- PHI encryption and access controls
- AI-specific guardrails and monitoring
Governance Templates
Classification schemes, access policies, lineage tracking, and retention rules for PHI
- PHI data classification
- Role-based access policies
- Audit and lineage tracking
- Retention and lifecycle rules
Risk Worksheets
Bias scoring, clinical risk catalogs, and mitigation planning worksheets
- Bias and fairness scoring
- Clinical risk catalogs
- Mitigation planning templates
- Pre-deployment checklists
Implementation Playbooks
End-to-end deployment and validation guides for regulated environments
- CI/CD for regulated infra
- Security validation steps
- Load and compliance testing
Glossary
Definitions of regulatory, AI/ML, cloud, and security terms used across this site
- HIPAA, FDA, GxP terms
- ML/LLM terminology
- Cloud infrastructure concepts
- Security and compliance terms